III. EXPERIENCES OF IWT MULTIMODAL TRANSPORT IN AND THE UNITED STATES

A. Europe

Multimodal container transport has grown tremendously in Europe over recent years. Thanks to an adapted infrastructure and adequate facilitation, traffic has been able to grow, often at a two-digit rate, throughout the last two decades.

1. Infrastructure

To permit this growth, a number of infrastructure elements have been created.

(a) Numerous container terminals

The first batch of container gantries was installed on the as far back as 1968, only two years after the first fully cellular ship arrived in Europe. However, in this new trade, which seemed to be vying for speed, it took some time for the deep-sea operators to consider the IWT option, which was purported to be slow. A number of container lines failed for lack of response by the market.

However, in the early 1980s, the cost savings brought about by fully dedicated river containerships was recognized, and local interests started building new container terminals, the farthest inland at Basel, Switzerland, 870 km from the sea. Today, there are some 40 container terminals along the banks of the Rhine. They serve the container transport on the Rhine and its seaports in the delta, mainly Rotterdam in the and Antwerp in .

Similarly, on the Rhone and Seine rivers in , services started in the 1990s, with eight inland terminals. In also, some terminals were built in other river basins, such as the Danube, Elbe and Weser. Each of these smaller basins serves a large seaport, such as Le Havre, Dunkirk and Marseille in France, and Hamburg and Bremen in Germany. Finally, in Belgium and the Netherlands, there are some 20 terminals servicing their respective seaports, Antwerp and Rotterdam.

-19- (b) Suitable waterway network

This container transport, however, is limited to routes where the necessary gauge can be found.

A great facility in is the free-flowing Rhine, unimpeded by locks and with ample vertical clearance under bridges, allowing four layers of containers up to Strasbourg in France, 700 km inland. It is on this river that major development has taken place, with more than 1 million containers crossing, for instance, the border between Germany and the Netherlands.

Canalized waterways, with so-called “European Gauge” locks (12 m wide and more than 110 m long), are also used for container traffic. The vertical clearance has often been critical, and most of the successful container services plying waterways allow three layers of containers. Some canals are even larger than the European Gauge, for instance the Scheldt- Rhine Canal between Rotterdam and Antwerp, where the locks are 24 m wide and the use of barges with four layers of containers is sometimes possible. The traffic on this particular waterway is of the same magnitude as on the Rhine, which shows the benefit that two close ports can derive from a strong waterway connection. However, there are examples of services plying with only two layers of containers, for instance in the northern part of France or in the Netherlands.

As is the case for vertical restrictions, there are instances where services are profitable even in narrower waterways, for instance in France and the Netherlands. There is a plan to develop the use of so-called Neo Kemp, which can squeeze a maximum number of containers in smaller canals. In addition, there are talks in France of developing a “Freycinet 2000”, plying in the smaller network with only one layer of containers and a capacity of 10 TEUs.

The reason behind these investment proposals is that the cost of barging is only a fraction of the total door-to-door transport cost, and that by bringing the containers closer to the market, the savings on the road leg are expected to be far greater than any augmentation of IWT cost.

Road transport, hampered by road congestion and overcrowded access to terminals, cannot guarantee delivery time in all cases. However, inland waterways are free from traffic congestion.

-20- (c) Adapted vessels

In Europe, the main concern is the vertical Figure III.1 clearance under bridges. Accordingly, all European designs but one are using hopper barges, in which the containers are lowered much in the same way as in a ship.

All started with normal bulk barges, which accommodated only three rows of containers, because of slanted sides, and three layers only, owing to stability considerations. A Europa II pushed barge, 11.4-m wide, could in that way accommodate 90 to 99 TEUs. The pusher had to use a telescopic wheelhouse (figure III.1) to be able to see above and in front of this high barge. Photo by courtesy of CFNR

Soon self-propelled barges were introduced, mainly by owner-operators. Quickly, it was found possible to squeeze in the hold of an 11.4-m wide barge four rows of containers, giving the possibility to stack four to five Figure III.2 layers and increase capacity up to 220 TEUs. Barges were further lengthened up to 110 m, thus carrying up to 280 TEUs (in five layers). Combined with a pushed barge like the one shown in figure III.1, such a unit loads up to 480 TEUs with five layers.

Finally, wider barges have been used, some of which are also cellular vessels (figure III.2). These vessels carry up to 480 TEUs. Figure III.3

There is also the Neo Kemp, mentioned earlier, as seen in figure III.3. Like some Pearl River Delta barges, it stacks barges across, and has its own crane. It holds 40 TEUs and can squeeze into a fine network of smaller waterways, bringing it closer to the clients.

-21- On the larger waterways, Figure III.4 pushed convoys of four barges can move up to 800 TEUs at the same time. Figure III.4 shows a barge convoy carrying 447 TEUs in a pushed three-barge convoy, with a total capacity of 607 TEUs, more than many sea-going feeder ships.

There appears to be no limit to what IWT can offer, such as the river shuttles between Antwerp and Rotterdam, which carry in a coordinated manner 650,000 TEUs per year, on behalf of four different barge operators.

2. Facilitation

There have been a number of areas in which container trade has obtained concessions not available to other goods carriage.

(a) No tonnage limitation

One of the strong points for IWT has been that, from the start, container traffic on European rivers has been exempt from limitations, imposed by governments or the profession, regarding the tonnage of barges. This has been specifically noticeable in France and in the Netherlands, where there are controls on size and overall capacity of fleet.

All European countries have load limitations for trucks to protect road pavement and the environment. The legislation passed by the EU restricts the maximum weight of tractor- trailers to 40 gross tonnes, except for combined transport operations and some multimodal transport operations in the vicinity of seaports. This load limitation is 28 tonnes in Switzerland.

The maximum payload on a normal 40-tonne European trailer is 26 to 28 tonnes, depending on the number of axles and the weight of the trailer. Assuming two containers are carried on the same trailer, on the same route to or from the same address, each container then can be 13 to 14 tonnes.

-22- However, the same road trailer can load up to 44 tonnes if in combined transport, which means more or less 2 tonnes more weight for each container. Combined transport, and not only by IWT, thus has a niche market with the 15 to 16 gross tonnes of payload, because it reduces the cost of the road leg, which normally accounts for about half of the total cost of a combined transport.

(b) Free rates

Another point is that the rates are also free, providing an added incentive for barge owners. Not that they necessarily have very good rates, but at least they are free to decide whether to run at these prices or not. On all other trades, official freight rates are imposed, and very often this translates into prices too high to attract customers, who then choose other modes of transport.

(c) ICD status

The trade really picks up when river terminals are given ICD status by the sea operators, allowing customers to pick up or return empty containers inland, instead of being obliged to do so at the seaport, thus saving one leg of the route. This also ensures flexibility in obtaining containers at short notice, and customers are therefore easily attracted to this system, which is less cumbersome and far cheaper than road operations.

This ICD status is neither a right nor a favour; it is actually brought about by the market itself as soon as an area has shown to be a good traffic area. Then, operators find it more convenient to leave the empty containers where they are, rather than bringing them back to the port, and shifting them back later to the same inland location to meet the demand.

(d) Expeditious customs clearance

Another positive point is the possibility to complete customs formalities inland. Thanks to a general movement in favour of facilitation this has been made easier every year and the number of inland customs outlets is rising everywhere. IWT has not been the only beneficiary, but often, inland ports have been the forerunners in attracting these bureaus. Lately, it has been agreed on the Seine River that all customs documentation remaining in the seaport can be done while the barge is under way, which transforms the barge into a moving warehouse, similar to, for example, the “virtual gate” seen in Hong Kong, China. Passage of

-23- the containers in the ports is thus very fast and easy, a further guarantee that operations will be smooth.

(e) Legislation in favour of intermodal transport

The road legs are normally kept to a minimum in combined transport, fixed by law in some countries, to less than 150 km from the terminal.

Thus, within 150 km of each combined transport terminal, including IWT terminals, heavy containers carried by road could be carried two on a chassis, while in direct transport they would have to be a standalone. The intricacy of this system will be used in the case study (see chapter VII).

Some European countries limit the time at which trucks can be used on main roads. They have also placed other restrictions on trucking. Germany has decided to charge tax for use of its roads. Switzerland will place an embargo on trucks in transit through the country from 2004. These measures force traditional road users to seek alternatives, such as inland waterways, rail and intermodal transport.

(f) Financial measures in favour of combined transport

Governments and the EU have consistently provided financial support to combined transport, especially since the United Nations Conference on Environment and Development, held at Rio de Janeiro, Brazil, from 3 to 14 June 1992, and the ensuing engagements to limit the greenhouse effect.

This support is aimed at directing traffic to rail and inland waterways instead of long- distance road transport, which is congestion-prone and less environment-friendly.

This has entailed grants for building combined transport terminals, especially tri- modal terminals (rail, water, road). Up to 50 per cent of the cost of studies and up to 25 per cent of final infrastructure costs could be provided in this manner.

The new Marco Polo Programme proposal continues in the same vein, with added emphasis on supporting the start-up of services. As reported in the EC document

-24- COM (2002)54 final,11 “The PACT Programme has provided evidence that setting up new intermodal freight services in Europe is fraught with risk. Community funding for start-up of new operations seems therefore still appropriate… Today, if market actors’ willingness to take risks is not stimulated beyond the traditional commercial incentives, traffic will stay on road”. Details of the current Marco Polo aid proposal are found in annex I.

This aid on a European scale is complemented on a national basis, such as the French or Netherlands programmes to that effect, covering up to 25 per cent of infrastructure, with a ceiling of around US$ 2 million.

Moreover, to facilitate the shift from long-distance operations by truck into combined transport operations, the dedicated trucks are exempt, or rather reimbursed, from a number of road taxes, representing up to a US$ 1,000 saving per year and per truck.

3. Resulting traffic

A detailed review of IWT container traffic in Europe has been published by AFTM,12 stating that, according to one of the scenarios prepared by the Rotterdam Port, container share in its global IWT traffic would treble before 2020. It estimates that 7 million TEUs, nearly 80 million tonnes of manufactured products, would then be carried on West European waterways.

Yet, this first comprehensive review of Western Europe container traffic also shows that the 1998 figure of 2.4 million TEUs was already equal to the “low” prognosis announced by the Rotterdam Port for 2010 only years ago. Then, 7 million TEUs in 2020 appeared to be a bare minimum.

Updates of this series confirm this trend: figures for 2001 showed that, despite problems brought about by recession and floods, IWT container traffic had reached 3.3 million TEUs in Western Europe.

11 , 2002, COM(2002)54 final, on the granting of community financial assistance to improve the environmental performance of the freight transport system (Brussels). 12 Association for Fluid Multimodal Traffic (AFMT), 29 February 2000, Navigation, Ports et Industries, Strasbourg (Paris).

-25- Table III.1. West European IWT container traffic

Total Rhine Total Non-Rhine Non-Rhine Year Grand total traffic Delta traffic French traffic German traffic 1987 327,766 229,000 – 12,000 568,766 1988 383,641 325,000 – 30,000 738,641 1989 372,275 444,000 – 44,673 860,948 1990 446,296 433,000 – 53,556 932,852 1991 498,227 523,000 – 59,787 1,081,014 1992 458,057 570,000 – 54,967 1,083,024 1993 546,431 657,500 – 65,572 1,269,503 1994 607,748 746,000 1,700 72,930 1,428,378 1995 795,454 802,000 10,122 73,000 1,680,576 1996 936,634 975,000 17,733 100,000 2,029,367 1997 982,891 1,093,000 21,323 89,000 2,186,214 1998 1,028,283 1,265,000 21,441 66,000 2,380,724 1999 1,084,359 1,507,000 36,628 84,000 2,711,987 2000 1,260,081 1,679,500 58,273 105,000 3,102,854 2001 1,300,992 1,771,000 71,308 139,000 3,282,300

Sources: AFTM estimates from information obtained from Rotterdam and Antwerp ports, DeStatis, VNF and AVV Transport Research Centre.

The corresponding curve (figure III.5) is even more striking, and demonstrates the ability of IWT to adjust to this new business, specially since this growth took place during the recession of the 1990s.

(a) Container transport in the Rhine delta

One of the most interesting elements of this evolution has been the dramatic increase of container transport in the Rhine Delta, be it domestic transport in the Netherlands or exchange between Belgium, France and the Netherlands. It equaled Rhine traffic around 1989, and surpassed it largely in 1997.

For instance, the short distance domestic transport in the Netherlands has bloomed, and now reaches every corner of the country. There are even exchanges from terminal to terminal, without touching any seaport. Most striking, however, has been the emergence of the traffic in and out of Antwerp, most of it interchange traffic between the two major container ports. The traffic over the Scheldt-Rhine Canal is now of the same order as the cross-border traffic on the Rhine. Thanks to the cooperation of the Dutch (AVV) Transport

-26- Research Centre and the port of Antwerp, an updated calculation reconciling various sources has been possible.

Figure III.5. West European IWT container traffic

3,500,000 TOTAL TRAFFIC

3,282,000

3,103,000 3,000,000

2,712,000

2,500,000 2,381,000

2,186,000

2,000,000 2,029,000 DELTA TRAFFIC 1 ,771 ,000 1,681,000

1,500,000 1,428,000

1,270,000 1 ,301 ,000 1,083,000 RHINE TRAFFIC

1,081,000 1,000,000 933,000 861,000 739,000

569,000 500,000 448,000 370,000 300,000 230,000 186,000 131,000 NON-RHINE GERMAN TRAFFIC 98,000 NON-RHINE FRENCH TRAFFIC 0 1975 1980 1985 1990 1995 2000 2005

(i) Antwerp

The Antwerp document series comprises on the one hand the part of Antwerp activity to and from the Rhine, and on the other the traffic exchanged with Rotterdam, which is not included in Rhine statistics. There is also some intraport activity, as well as transport from Antwerp towards the rest of Belgium and the northern part of France.

-27- The curve in figure III.6 is extremely dynamic, and the port of Antwerp anticipates that it will remain so for the years to come, expecting that IWT might eventually replace road transport as the main hinterland carrier. The prognosis is for IWT to achieve this by the year 2010, with some 3.6 million TEUs.

Figure III.6. IWT container traffic in Antwerp

1,800,000 TOTAL 1,612,000 1,600,000 1,523,000

1,400,000 1,303,000 1,200,000 1,202,000

1,000,000 1,011,000

in TEU in 840,000 800,000 Antwerp-Rotterdam 674,000 707,000 600,000 580,000 Antwerp-Rhine 457,000 400,000 393,000 457,000 293,000 239,000 200,000 175,000 Towards Belgium and 120,000 North 165,000 68,000 5,000 0 1970 1975 1980 1985 1990 1995 2000 2005

Source: AFTM estimates, based on data by the AVV Transport Research Centre and the port of Antwerp.

The growing strength of IWT in container transport in Antwerp and the modal split has evolved as shown in table III.2.

Table III.2. Port of Antwerp modal split of hinterland distribution of containers (Without sea-sea trans-shipment, in percentage)

Mode 1995 1996 1997 1998 1999 2000 2001 Change 95/01 River 22.7 24.3 27.1 27.6 27.9 29.3 29.9 +7.2 Rail 5.2 6.2 7.1 7.8 9.3 10.1 8.8 +3.6 Road 72.1 69.5 65.8 64.6 62.8 60.6 61.3 -10.8 Total 100.0 100.0 100.0 100.0 100.0 100.0 100.0 –

Source: Regular survey by AGHA-SEA, 2002.

-28- It can be observed that the strength of IWT container traffic between Antwerp and Rotterdam casts a different light on the respective roles of these two as transit ports for containerized goods, Antwerp attracts containers from its hinterland as well as from Germany, while Rotterdam sends these overseas.

(ii) Delta series

Actually, to ensure that the traffic from Antwerp to the Rhine is not counted twice, another series is reported and is shown in figure III.5 and table III.1. It is necessary to consolidate all containers moving in the delta that are not included in the Rhine traffic, taken as the traffic upstream of the German border. This Delta series comprises the Rotterdam- Antwerp interchange, Rotterdam to Belgium or Northern France, Antwerp to the same destinations, and internal Netherlands traffic, including from Rotterdam.

Table III.3. Delta IWT container traffic Antwerp- Netherlands Intraport Belgium-North- Year Total delta traffic Rotterdam traffic domestic traffic traffic Pas-de-Calais 1987 125,000 40,000 60,000 4,000 229,000 1988 170,000 60,000 80,000 15,000 325,000 1989 195,000 130,000 100,000 19,000 444,000 1990 190,000 120,000 100,000 23,000 433,000 1991 275,000 120,000 100,000 28,000 523,000 1992 330,000 130,000 80,000 30,000 570,000 1993 400,000 140,000 85,000 32,500 657,500 1994 431,000 195,000 85,000 35,000 746,000 1995 451,000 210,000 85,000 56,000 802,000 1996 550,000 280,000 85,000 60,000 975,000 1997 602,000 320,000 83,000 88,000 1,093,000 1998 680,000 316,000 162,000 107,000 1,265,000 1999 647,000 526,000 209,000 125,000 1,507,000 2000 713,500 551,000 241,500 173,500 1,679,500 2001 707,000 576,000 283,000 205,000 1,771,000 Sources: For 1993 and 1996, Brolsma, AVV Transport Research Centre, PIANC Bulletin, June 1997; for other years, AFTM estimates based on data from AVV Transport Research Centre and port of Antwerp.

Figure III.7. Container traffic at the G erm an-N etherlands border (b) Container transport on the Rhine 1,500,000

1,300,000 1,285,000 Another interesting item to observe is 1,130,000 1,100,000 the passage at the German-Netherlands border, 1,023,000 900,000 904,000 TEU 832,000 as shown in figure III.7 (Emmerich-Lobith). 778,000 700,000 657,000 An estimate in TEU is available 640,000 500,000

300,000 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002

-29- since 1994.13 After a faltering start, its growth has been sustained, and now reaches nearly 1.3 million TEUs.

The so-called “Rhine Traffic” includes some other flows, internal to the Rhine basin, and thus the figures vary slightly from those at the border.

(c) Non-Rhine German traffic

Figures for this are less well documented and refer to the Weser, Elbe, Ems and Danube basins, once figures for traffic related to the Rhine delta are taken out. Before 1996, they were counted as 12 per cent of Rhine traffic.

(d) Non-Rhine French traffic

Traffic on the Rhone started at the end of the 1970s, but volumes rarely surpassed 6,000 TEUs. Since, services have been suspended two or three times, statistics in table III.4 show only figures available since the reopening of the Deltabox line in 1994.

On the Seine, traffic actually began only at the end of 1994 with LOGISEINE, a provider of river-road transport services on the Seine although earlier attempts were made in the 1980s.

In the North-Pas-de-Calais region, a service linking Dunkirk with its hinterland started in 1999. That same year, a domestic container service, carrying urban waste to a disposal ground, started on a strong footing. The other services out of the region are counted as Rhine Delta traffic, since they connect with the two main seaports there. The non-Rhine French traffic series is shown in table III.4.

Table III.4. Non-Rhine French traffic

Dunkirk-Lille- Non-maritime Year Seine Rhone Total Valenciennes containers 1994 500 1,200 – – 1,700 1995 5,811 4,311 – – 10,122 1996 11,433 6,300 – – 17,733 1997 16,598 4,725 – – 21,323 1998 14,927 6,514 – – 21,441 1999 21,000 3,388 1,240 9,298 36,628 2000 32,700 2,987 3,515 19,071 58,273 2001 38,400 8,635 5,725 18,548 71,308 Source: AFTM estimates based on companies’ data and VNF statistics.

The rate of growth in 2001 was 22 per cent and was expected to continue.

13 Central Commission for Navigation on the Rhine (CCNR), 2001, Economic evolution of Rhine Navigation, Statistics 2001. -30- (e) Danube

Non-German Danube traffic represents thousands of TEUs. No reliable series are available. Figures of 100,000 TEUs in 1990 and 160,000 TEUs in 1993 are proposed, but seem to include all modes of transport. For lack of more precise indications, Danube traffic will not be included in the West European total.

B. United States

Very little container traffic is found on main inland waterways although overall IWT traffic is quite high in the United States.

Due to the general orientation of the river network, which is north-south, and which therefore does not coincide with the container movements, generally going east-west, no stable container traffic has developed on the Mississippi River. Furthermore, rail with its double stack trains, has become strong competition.

However, since July 2002, the port of Baton Rouge on the Mississippi River has been operating a weekly regular line between Louisiana and Texas, which has the advantage of running in an east-west rather than the dominant north-south direction. Traffic was expected to reach 40,000 TEUs in 2002.

One place where a regular container line has been able to truly prosper is the Columbia-Snake Waterway, 700 km long, linking from East to West the State of Idaho to the sea in Portland, Oregon. It has even initiated traffic in some very specific commodities for exports, such as frozen french fries or cubes of compressed hay. Traffic started in 1977 at a rather low level (7,000 TEUs), because Portland is a somewhat small container port (less than 300,000 TEUs). Traffic reached a peak of 91,000 TEUs in 2000. Figure III.8 Worth mentioning is that during 2000 and 2001, intermodal transport carried more than 60 per cent of the containers at the port of Portland, with 31 per cent of total container traffic by IWT. This is a percentage unrivalled anywhere in the world, except in the inland port of Basel, Switzerland.

-31- The types of barges used are either deck barges as shown in figure III.8 or hopper barges to better protect the containers.

The two systems coexist. In Figure III.9 figure III.9, an 80-TEU hopper reefer barge can be seen, with a 40-foot reefer container being handled, which started operations in 1991. Of the containers carried, 75 per cent are 40-foot containers.

Further, most barges in the United States are pushed barges, figure III.10 shows a barge being Figure III.10 pushed across a reservoir.

Hay, a very cheap commodity and as such very transport cost- sensitive, is carried over a long distance, 600 km. Barging costs are typically 20 to 50 per cent lower than rail or trucking. There is little doubt that, had this less expensive alternative not been available, this never would have materialized.

Interesting to note is that IWT increases the hinterland of the port of Portland, faced with competition from the far bigger ports of Seattle and Tacoma. Similarly, as explained below, IWT service increases the hinterland of ports in Europe, such as Le Havre, faced with competition from Antwerp.

In all zones where parity exists between Seattle and Portland trucking, the superior sea-service frequency of Seattle should normally weigh heavily in its favour. Instead, the 20 to 50 per cent cost reduction of barging compared with trucking secures the loads to Portland. A gain of up to US$ 350 per TEU is reported in some cases, compared with trucking to Seattle.

The service provided by the barge operators has been tailored to the timing of the sea liner calls. There are two services per week and per river operator, each reaching the port in

-32- time for the weekly call of a specific Figure III.11 sea liner. Most calls are on Mondays and Fridays, and on those days barges line up to wait for the sea- going vessels (figure III.11), providing some 250 TEUs per ship call.

In 2000, barging rates were around US$ 0.17 per TEU-km, and US$ 0.24 per 40-foot container per km. To this must be added inland port handling charges, and the occasional trucking. These costs are quite similar to those found in Europe.

Recently, the US Department of Transportation funded studies on container-on-barge transport in the Mississippi River basin, notably a feasibility assessment of truck-barge intermodal freight transport and empty container management for container-on-board transport. This is one of the key issues in order to break new ground and demonstrate the relevance of IWT for the transport of containers, even in a highly competitive environment.

The port of New York has also come up with the concept of a port inland distribution network (PIDN), and envisions a network of inland container terminals, linked to the port by dedicated rail, barge or tandem trailer-truck shuttle services. The PIDN is expected to reduce congestion, improve efficiency and reduce the cost of inland transport by 20 per cent.

It is hoped that these studies will lead to new container-on-barge developments taking place in the United States.

-33-